claude-mpm 4.16.0__py3-none-any.whl → 4.25.10__py3-none-any.whl

claude_mpm/services/mcp_gateway/tools/kuzu_memory_service.py

@@ -199,7 +199,9 @@ class KuzuMemoryService(BaseToolAdapter):
         try:
             if action == "store":
                 result = await self.store_memory(
-                    params.get("content"), params.get("tags"), {}  # metadata
+                    params.get("content"),
+                    params.get("tags"),
+                    {},  # metadata
                 )
             elif action == "recall":
                 result = await self.recall_memories(
@@ -213,7 +215,9 @@ class KuzuMemoryService(BaseToolAdapter):
                 )
             elif action == "context":
                 result = await self.get_context(
-                    params.get("query", ""), 2, True  # depth  # include_related
+                    params.get("query", ""),
+                    2,
+                    True,  # depth  # include_related
                 )
             else:
                 return MCPToolResult(success=False, error=f"Unknown action: {action}")

claude_mpm/services/memory/failure_tracker.py

@@ -29,6 +29,7 @@ Example flow:
 
 import logging
 import re
+import threading
 from dataclasses import dataclass, field
 from datetime import datetime, timezone
 from typing import Dict, List, Optional, Tuple
@@ -536,6 +537,7 @@ class FailureTracker:
 
 # Singleton instance for session-level tracking
 _tracker_instance: Optional[FailureTracker] = None
+_tracker_lock = threading.Lock()
 
 
 def get_failure_tracker() -> FailureTracker:
@@ -544,13 +546,23 @@ def get_failure_tracker() -> FailureTracker:
     WHY: Session-level tracking requires a singleton to maintain state
     across multiple hook invocations during the same session.
 
+    Thread-safe implementation using double-checked locking pattern to
+    prevent race conditions during concurrent initialization.
+
     Returns:
         The FailureTracker singleton instance
     """
     global _tracker_instance
-    if _tracker_instance is None:
-        _tracker_instance = FailureTracker()
-    return _tracker_instance
+
+    # Fast path - check without lock
+    if _tracker_instance is not None:
+        return _tracker_instance
+
+    # Slow path - acquire lock and double-check
+    with _tracker_lock:
+        if _tracker_instance is None:
+            _tracker_instance = FailureTracker()
+        return _tracker_instance
 
 
 def reset_failure_tracker() -> None:
@@ -558,6 +570,9 @@ def reset_failure_tracker() -> None:
 
     WHY: Tests need to reset state between runs. Also useful for
     explicitly starting a new tracking session.
+
+    Thread-safe implementation ensures proper cleanup.
     """
     global _tracker_instance
-    _tracker_instance = None
+    with _tracker_lock:
+        _tracker_instance = None

claude_mpm/services/memory/optimizer.py

@@ -618,7 +618,7 @@ class MemoryOptimizer(LoggerMixin):
                     )
 
                 # Check for similar items
-                consolidated, consolidated_count = self._consolidate_similar_items(
+                _consolidated, consolidated_count = self._consolidate_similar_items(
                     unique_points
                 )
                 if consolidated_count > 0:

claude_mpm/services/model/model_router.py

@@ -266,7 +266,7 @@ class ModelRouter(BaseService, IModelRouter):
 
         # Route based on strategy
         if self.strategy == RoutingStrategy.CLAUDE_ONLY:
-            return await self._route_to_claude(content, task, model, **kwargs)
+            return await self._route_to_claude(content, task, model=model, **kwargs)
 
         if self.strategy in (
             RoutingStrategy.OLLAMA_ONLY,
@@ -275,13 +275,13 @@ class ModelRouter(BaseService, IModelRouter):
             return await self._route_to_ollama(
                 content,
                 task,
-                model,
+                model=model,
                 require_success=True,
                 **kwargs,
             )
 
         # AUTO strategy
-        return await self._route_auto(content, task, model, **kwargs)
+        return await self._route_auto(content, task, model=model, **kwargs)
 
     async def _route_auto(
         self,
@@ -308,7 +308,7 @@ class ModelRouter(BaseService, IModelRouter):
             response = await self._route_to_ollama(
                 content,
                 task,
-                model,
+                model=model,
                 require_success=False,
                 **kwargs,
             )
@@ -325,7 +325,7 @@ class ModelRouter(BaseService, IModelRouter):
         # Ollama unavailable or failed - fallback to Claude
         if self.fallback_enabled:
             self.log_info("Falling back to Claude")
-            return await self._route_to_claude(content, task, model, **kwargs)
+            return await self._route_to_claude(content, task, model=model, **kwargs)
         return self._create_error_response(
             task,
             model,
@@ -356,8 +356,7 @@ class ModelRouter(BaseService, IModelRouter):
         if require_success and not await self.ollama_provider.is_available():
             if self.strategy == RoutingStrategy.PRIVACY_FIRST:
                 error_msg = (
-                    "Ollama not available. Privacy mode enabled - "
-                    "not sending to cloud."
+                    "Ollama not available. Privacy mode enabled - not sending to cloud."
                 )
             else:
                 error_msg = "Ollama not available and required by configuration"
@@ -368,7 +367,7 @@ class ModelRouter(BaseService, IModelRouter):
         self._route_count["ollama"] += 1
 
         return await self.ollama_provider.analyze_content(
-            content, task, model, **kwargs
+            content, task, model=model, **kwargs
         )
 
     async def _route_to_claude(
@@ -394,7 +393,7 @@ class ModelRouter(BaseService, IModelRouter):
         self._route_count["claude"] += 1
 
         return await self.claude_provider.analyze_content(
-            content, task, model, **kwargs
+            content, task, model=model, **kwargs
         )
 
     def _create_error_response(

claude_mpm/services/monitor/daemon.py

@@ -479,7 +479,7 @@ class UnifiedMonitorDaemon:
             except OSError:
                 if i < 9:
                     self.logger.debug(
-                        f"Port {self.port} still in use, waiting... ({i+1}/10)"
+                        f"Port {self.port} still in use, waiting... ({i + 1}/10)"
                     )
                     time.sleep(1)
                 else:

claude_mpm/services/monitor/server.py

@@ -203,8 +203,8 @@ class UnifiedMonitorServer:
             # Create Socket.IO server with proper ping configuration
             self.sio = socketio.AsyncServer(
                 cors_allowed_origins="*",
-                logger=False,
-                engineio_logger=False,
+                logger=True,  # Enable to see Socket.IO events and connection lifecycle
+                engineio_logger=True,  # Enable to see Engine.IO protocol handshake details
                 ping_interval=30,  # 30 seconds ping interval (matches client expectation)
                 ping_timeout=60,  # 60 seconds ping timeout (generous for stability)
             )

claude_mpm/services/native_agent_converter.py

@@ -0,0 +1,356 @@
+"""Native Agent Converter Service
+
+Converts MPM agent JSON definitions to Claude Code --agents flag format.
+
+WHY: Claude Code 1.0.83+ supports native --agents flag for dynamic agent definition.
+This allows agents to be passed directly via CLI instead of deployed to .claude/agents/.
+
+DESIGN: Converts MPM agent schema → Claude native schema with field mappings:
+- description → description (agent selection hint)
+- system_instructions → prompt (agent behavior)
+- allowed_tools → tools (tool limitations)
+- model_tier → model (model selection)
+
+USAGE:
+    converter = NativeAgentConverter()
+    agents_flag = converter.build_agents_flag(agent_configs)
+    # Returns: --agents '{"agent-name": {...}, ...}'
+"""
+
+import json
+from pathlib import Path
+from typing import Any, Dict, List, Optional
+
+from claude_mpm.core.logging_config import get_logger
+
+
+class NativeAgentConverter:
+    """Converts MPM agent configurations to Claude Code --agents format."""
+
+    # Map MPM model tiers to Claude model names
+    MODEL_TIER_MAP = {
+        "opus": "opus",
+        "sonnet": "sonnet",
+        "haiku": "haiku",
+        "claude-3-opus": "opus",
+        "claude-3-sonnet": "sonnet",
+        "claude-3-haiku": "haiku",
+        "claude-3.5-sonnet": "sonnet",
+        "claude-4-sonnet": "sonnet",
+        "claude-4-opus": "opus",
+    }
+
+    # Map MPM tool names to Claude tool names
+    TOOL_NAME_MAP = {
+        "Read": "Read",
+        "Write": "Write",
+        "Edit": "Edit",
+        "MultiEdit": "MultiEdit",
+        "Bash": "Bash",
+        "Grep": "Grep",
+        "Glob": "Glob",
+        "LS": "LS",
+        "WebSearch": "WebSearch",
+        "WebFetch": "WebFetch",
+        "TodoWrite": "TodoWrite",
+        "NotebookEdit": "NotebookEdit",
+        "BashOutput": "BashOutput",
+        "KillShell": "KillShell",
+        "AskUserQuestion": "AskUserQuestion",
+    }
+
+    def __init__(self):
+        """Initialize the native agent converter."""
+        self.logger = get_logger(__name__)
+
+    def convert_mpm_agent_to_native(
+        self, agent_config: Dict[str, Any], agent_id: Optional[str] = None
+    ) -> Dict[str, Any]:
+        """Convert a single MPM agent config to Claude native format.
+
+        Args:
+            agent_config: MPM agent JSON configuration
+            agent_id: Optional agent ID (falls back to agent_config['agent_id'])
+
+        Returns:
+            Dict with Claude native agent format:
+            {
+                "description": "...",
+                "prompt": "...",
+                "tools": [...],
+                "model": "sonnet"
+            }
+        """
+        try:
+            # Extract agent ID
+            if not agent_id:
+                agent_id = agent_config.get("agent_id") or agent_config.get("name", "")
+
+            # Extract description (for agent selection)
+            description = agent_config.get("description", "")
+            if not description and "metadata" in agent_config:
+                description = agent_config["metadata"].get("description", "")
+
+            # Build prompt from instructions and BASE_*.md reference
+            prompt = self._build_agent_prompt(agent_config)
+
+            # Extract and map tools
+            tools = self._extract_and_map_tools(agent_config)
+
+            # Map model tier
+            model = self._map_model_tier(agent_config)
+
+            native_config = {
+                "description": description,
+                "prompt": prompt,
+                "tools": tools,
+                "model": model,
+            }
+
+            self.logger.debug(f"Converted agent '{agent_id}' to native format")
+            return native_config
+
+        except Exception as e:
+            self.logger.error(f"Error converting agent {agent_id}: {e}")
+            # Return minimal valid config as fallback
+            return {
+                "description": f"Agent {agent_id}",
+                "prompt": agent_config.get("instructions", ""),
+                "tools": [],
+                "model": "sonnet",
+            }
+
+    def _build_agent_prompt(self, agent_config: Dict[str, Any]) -> str:
+        """Build agent prompt from instructions and BASE_*.md reference.
+
+        OPTIMIZATION: Keep prompts concise for CLI argument length limits.
+        The BASE_*.md files contain full instructions, so we only need:
+        1. Reference to BASE file
+        2. Brief specialization note
+
+        Args:
+            agent_config: MPM agent configuration
+
+        Returns:
+            Concise prompt string
+        """
+        prompt_parts = []
+
+        # Add base instructions reference if available (most important)
+        if "knowledge" in agent_config:
+            base_file = agent_config["knowledge"].get("base_instructions_file")
+            if base_file:
+                prompt_parts.append(f"Follow {base_file} for all protocols.")
+
+        # Add main instructions (keep brief)
+        instructions = agent_config.get("instructions", "")
+        if instructions:
+            # Limit instruction length to avoid bloat
+            if len(instructions) > 300:
+                instructions = instructions[:300] + "..."
+            prompt_parts.append(instructions)
+
+        # Skip domain expertise and best practices for CLI mode
+        # These are already in BASE_*.md files referenced above
+        # Adding them here just bloats the JSON unnecessarily
+
+        return "\n".join(str(part) for part in prompt_parts if part)
+
+    def _extract_and_map_tools(self, agent_config: Dict[str, Any]) -> List[str]:
+        """Extract and map tools from MPM config to Claude tool names.
+
+        Args:
+            agent_config: MPM agent configuration
+
+        Returns:
+            List of Claude tool names
+        """
+        tools = []
+
+        # Check capabilities.tools
+        if "capabilities" in agent_config and "tools" in agent_config["capabilities"]:
+            mpm_tools = agent_config["capabilities"]["tools"]
+            for tool in mpm_tools:
+                mapped_tool = self.TOOL_NAME_MAP.get(tool, tool)
+                if mapped_tool not in tools:
+                    tools.append(mapped_tool)
+
+        # If no tools specified, provide reasonable defaults
+        if not tools:
+            tools = ["Read", "Write", "Edit", "Bash", "Grep", "Glob"]
+
+        return tools
+
+    def _map_model_tier(self, agent_config: Dict[str, Any]) -> str:
+        """Map MPM model tier to Claude model name.
+
+        Args:
+            agent_config: MPM agent configuration
+
+        Returns:
+            Claude model name (opus, sonnet, haiku)
+        """
+        # Check capabilities.model
+        if "capabilities" in agent_config and "model" in agent_config["capabilities"]:
+            model_tier = agent_config["capabilities"]["model"]
+            return self.MODEL_TIER_MAP.get(model_tier.lower(), "sonnet")
+
+        # Check metadata.model_preference
+        if (
+            "metadata" in agent_config
+            and "model_preference" in agent_config["metadata"]
+        ):
+            model_tier = agent_config["metadata"]["model_preference"]
+            return self.MODEL_TIER_MAP.get(model_tier.lower(), "sonnet")
+
+        # Default to sonnet
+        return "sonnet"
+
+    def generate_agents_json(self, agents: List[Dict[str, Any]]) -> str:
+        """Generate complete --agents JSON string from list of agent configs.
+
+        Args:
+            agents: List of MPM agent configurations
+
+        Returns:
+            JSON string for --agents flag
+        """
+        native_agents = {}
+
+        for agent_config in agents:
+            agent_id = agent_config.get("agent_id") or agent_config.get("name", "")
+            if not agent_id:
+                self.logger.warning("Skipping agent without ID")
+                continue
+
+            # Skip PM agent (main Claude instance)
+            if agent_id.lower() in ["pm", "project_manager"]:
+                self.logger.debug(f"Skipping PM agent: {agent_id}")
+                continue
+
+            native_config = self.convert_mpm_agent_to_native(agent_config, agent_id)
+            native_agents[agent_id] = native_config
+
+        return json.dumps(native_agents, separators=(",", ":"))
+
+    def build_agents_flag(
+        self, agents: List[Dict[str, Any]], escape_for_shell: bool = True
+    ) -> str:
+        """Build complete --agents flag for CLI.
+
+        Args:
+            agents: List of MPM agent configurations
+            escape_for_shell: Whether to escape JSON for shell
+
+        Returns:
+            Complete flag string: --agents '{"agent1": {...}, ...}'
+        """
+        agents_json = self.generate_agents_json(agents)
+
+        # Check length (Claude CLI has argument length limits)
+        if len(agents_json) > 50000:  # Conservative limit
+            self.logger.warning(
+                f"Agents JSON is very large ({len(agents_json)} chars). "
+                "Consider using file-based deployment."
+            )
+
+        if escape_for_shell:
+            # Escape for shell - wrap in single quotes
+            return f"--agents '{agents_json}'"
+
+        return f"--agents {agents_json}"
+
+    def load_agents_from_templates(
+        self, templates_dir: Optional[Path] = None
+    ) -> List[Dict[str, Any]]:
+        """Load all agent configs from templates directory.
+
+        Args:
+            templates_dir: Path to templates directory (defaults to MPM agents)
+
+        Returns:
+            List of agent configurations
+        """
+        if not templates_dir:
+            # Default to MPM agents directory
+            mpm_package_dir = Path(__file__).parent.parent / "agents" / "templates"
+            templates_dir = mpm_package_dir
+
+        if not templates_dir.exists():
+            self.logger.warning(f"Templates directory not found: {templates_dir}")
+            return []
+
+        agents = []
+        json_files = list(templates_dir.glob("*.json"))
+
+        self.logger.info(
+            f"Loading {len(json_files)} agent templates from {templates_dir}"
+        )
+
+        for json_file in json_files:
+            try:
+                # Skip base_agent.json
+                if json_file.stem == "base_agent":
+                    continue
+
+                agent_config = json.loads(json_file.read_text())
+                agents.append(agent_config)
+                self.logger.debug(f"Loaded agent: {json_file.stem}")
+
+            except Exception as e:
+                self.logger.error(f"Error loading agent {json_file.name}: {e}")
+                continue
+
+        return agents
+
+    def estimate_json_size(self, agents: List[Dict[str, Any]]) -> int:
+        """Estimate the size of --agents JSON output.
+
+        Args:
+            agents: List of agent configurations
+
+        Returns:
+            Estimated size in bytes
+        """
+        agents_json = self.generate_agents_json(agents)
+        return len(agents_json.encode("utf-8"))
+
+    def get_conversion_summary(self, agents: List[Dict[str, Any]]) -> Dict[str, Any]:
+        """Get summary of agent conversion for reporting.
+
+        Args:
+            agents: List of agent configurations
+
+        Returns:
+            Summary dict with counts and size info
+        """
+        native_agents = {}
+        model_counts = {"opus": 0, "sonnet": 0, "haiku": 0}
+        tool_usage = {}
+
+        for agent_config in agents:
+            agent_id = agent_config.get("agent_id") or agent_config.get("name", "")
+            if not agent_id or agent_id.lower() in ["pm", "project_manager"]:
+                continue
+
+            native_config = self.convert_mpm_agent_to_native(agent_config, agent_id)
+            native_agents[agent_id] = native_config
+
+            # Count models
+            model = native_config.get("model", "sonnet")
+            model_counts[model] = model_counts.get(model, 0) + 1
+
+            # Count tools
+            for tool in native_config.get("tools", []):
+                tool_usage[tool] = tool_usage.get(tool, 0) + 1
+
+        json_size = len(json.dumps(native_agents, separators=(",", ":")))
+
+        return {
+            "total_agents": len(native_agents),
+            "json_size": json_size,
+            "json_size_kb": round(json_size / 1024, 2),
+            "model_distribution": model_counts,
+            "tool_usage": tool_usage,
+            "agents": list(native_agents.keys()),
+        }

claude_mpm/services/port_manager.py

@@ -431,7 +431,7 @@ class PortManager:
                             return port
 
         self.logger.error(
-            f"No available ports in range {self.PORT_RANGE.start}-{self.PORT_RANGE.stop-1}"
+            f"No available ports in range {self.PORT_RANGE.start}-{self.PORT_RANGE.stop - 1}"
         )
         return None
 

claude_mpm/services/project/documentation_manager.py

@@ -369,7 +369,8 @@ class DocumentationManager:
         sorted_sections = sorted(
             sections.items(),
             key=lambda x: self.SECTION_PRIORITY.get(
-                self._get_section_key(x[0]), 50  # Default priority
+                self._get_section_key(x[0]),
+                50,  # Default priority
             ),
             reverse=True,  # Higher priority first
         )

claude_mpm/services/project/toolchain_analyzer.py

@@ -318,7 +318,9 @@ class ToolchainAnalyzerService(BaseService, IToolchainAnalyzer):
                     else (
                         3
                         if f.confidence == ConfidenceLevel.MEDIUM
-                        else 2 if f.confidence == ConfidenceLevel.LOW else 1
+                        else 2
+                        if f.confidence == ConfidenceLevel.LOW
+                        else 1
                     )
                 ),
                 f.popularity_score,

claude_mpm/services/runner_configuration_service.py

@@ -168,6 +168,7 @@ class RunnerConfigurationService(BaseService, RunnerConfigurationInterface):
             "launch_method": kwargs.get("launch_method", "exec"),
             "enable_websocket": kwargs.get("enable_websocket", False),
             "websocket_port": kwargs.get("websocket_port", 8765),
+            "use_native_agents": kwargs.get("use_native_agents", False),
         }
 
         # Initialize main configuration using ConfigLoader